The other day, a member of our sales team shared this Twitter thread comparing the usability of baby clothes to enterprise software.

This takedown is both funny and true, speaking as a parent and as an enterprise software co-founder and user. The really insightful bit is where the author talks about how value alignment at most enterprise software companies is broken. He correctly points out that the level of indirection between decision makers and users creates incentives for software companies to load up on features of dubious value and confuse the overall user base, which leads to non-adoption and dissatisfaction.

When we started data.world, I was inspired by companies that thoughtfully combine data-driven agile product development with design thinking.  While these companies typically rule consumer segments (e.g., Airbnb and Netflix), they do also exist in the enterprise space (e.g., Slack). We have an incredibly ambitious mission at data.world, to create the most meaningful, collaborative, and abundant data resource in the world. That involves democratizing working with data and publishing analysis at a rate people haven’t seen before. I drew from the difficulties I experienced organizing the data processes and knowledge management at Homeaway.com (where I previously worked, an Airbnb competitor now known as VRBO).  In order to make this vision a reality, we knew that design would need to have a seat near the head of the table. We hired two stellar UX designers. While many startups see design as either a luxury or an afterthought, it was an imperative at the inception of data.world.

In the ensuing three-plus years, we’ve built the world’s largest collaborative open data catalog and community.  We’ve worked with and interviewed hundreds of analysts, data scientists, and engineers, and we realized that to have a vibrant ecosystem, our product needed to be inclusive of less-technical subject matter experts who could describe the data.  In addition to the user research we do, we also collect an astounding amount of data about the usage patterns of our application, and we A/B test alternatives to see how they impact engagement and collaboration. This is the sort of thing that most business software companies just can’t do because they’re not cloud-based SaaS, so they don’t have a free tier that allows them to test designs with hundreds of thousands of people.

That brings us to the fall of 2018. While we were really happy with the work we were doing to realize our mission as a B-Corp and the progress we were making driving data literacy and inclusion forward, more and more enterprises approached us and wanted to replicate our tools and create internal data communities of their own. They wanted a data catalog for enterprises that did something a little different.  But enterprises buy data tools differently than they buy Slack or Github. Since enterprises value data so highly and protect it accordingly, it’s tough to execute a bottoms-up sales approach. There are security reviews, and people won’t sneak shadow IT into the land of analytics.  Because of this, many of the patterns that Arvind points out in his Twitter thread started to emerge: the feature checklists, long RFPs, differences between buyers and users, etc.

One way we ensure user satisfaction with data.world is another measure called C-RAD.  C-RAD is short for “Customer Real Active Day.” It’s basically the number of unique paid customer return visits to the data.world platform where the user has a meaningful session in the system. What constitutes a “meaningful” session is proprietary, but it’s based on the collaborative actions we observed in from the community working on our massive, free open data catalog. We expect this number to grow at customers on a quarter-over-quarter basis and we incentivize our entire company on C-RAD growth. This way everyone is aligned around growing actual usage, not just dollars, for each account. It’s critical to our philosophy that a customer that succeeds in rolling out data.world should succeed economically. That creates radical alignment and teamwork here, which helps us create productive data citizens within customer companies.  It has also helped us avoid the pitfalls mentioned in Arvind’s twitter thread, and it keeps the company focused on the success of end-users, not just the feature list creators.

Recently, we were lucky enough to be included in Gartner’s Magic Quadrant for Metadata Management.  For a company that recently made the jump to enterprise selling, that’s a massive accomplishment. Gartner put us in the “Niche Players” quadrant and wrote some nice things about us, especially on our ease of use and collaborative features. I’ll gladly be niche when that’s called out as our strength! Most importantly, our customers are saying great things about us in Gartner Peer Insights as well, and they too highlight our ease of use and dedication to our users.

We’re extremely proud of being a different kind of enterprise software company.  We truly put our users and community members at the center of everything we do. We’re so dedicated to it we’re willing to align our pay incentives around it to make sure every single employee works toward platform adoption goals. We fundamentally believe that our product has to offer a great experience to get the kind of widespread adoption that can actually make organizations data-driven.  That’s pretty radical (pun intended).

Want to learn more about the metadata management space? Download a complimentary copy of Gartner’s Magic Quadrant for Metadata Management here.